headers
Ignatios Souvatzis | 2 Feb 2010 15:46
Picon

ATI 0x1002/0x9498: how?

I'm using modular Xorg from pgsrc-2009Q4.

The card is: Radion 4650 

kolmogorov# pcictl /dev/pci1 list 
001:00:0: ATI Technologies product 0x9498 (VGA display)

outputs are : VGA-0 + DVI-0 + HDMI-0 (they've sent us that instead of
the double-DVI I wanted).

It seems to use xf86-video-radeon:

(II) Loading /usr/pkg/lib/xorg/modules/extensions//libdbe.so
(II) Loading /usr/pkg/lib/xorg/modules/extensions//libdri.so
(II) Loading /usr/pkg/lib/xorg/modules/extensions//libdri2.so
(II) Loading /usr/pkg/lib/xorg/modules/extensions//libextmod.so
(II) Loading /usr/pkg/lib/xorg/modules/extensions//libglx.so
(II) Loading /usr/pkg/lib/xorg/modules/drivers//radeon_drv.so
(II) Loading /usr/pkg/lib/xorg/modules/input//mouse_drv.so
(II) Loading /usr/pkg/lib/xorg/modules/input//kbd_drv.so
(II) Loading /usr/pkg/lib/xorg/modules//libvgahw.so
(II) Loading /usr/pkg/lib/xorg/modules//libint10.so
(II) Loading /usr/pkg/lib/xorg/modules//libshadow.so
(II) Loading /usr/pkg/lib/xorg/modules//libfb.so
(II) AIGLX: Loaded and initialized /usr/pkg/lib/dri/swrast_dri.so

I seem to have two problems:

a) I can't use xrandr to rotate DVI-0
b) I can't make HDMI-0 to use a useful resolution; only 640x480 seems
(Continue reading)

Permalink | Reply | 1 comment |
headers
Rafal Boni | 4 Feb 2010 20:54
Picon
Favicon

Re: ATI 0x1002/0x9498: how?

On Tue, Feb 02, 2010 at 03:46:10PM +0100, Ignatios Souvatzis wrote:
> I'm using modular Xorg from pgsrc-2009Q4.
> 
> The card is: Radion 4650 
> 
> kolmogorov# pcictl /dev/pci1 list 
> 001:00:0: ATI Technologies product 0x9498 (VGA display)
> 
> outputs are : VGA-0 + DVI-0 + HDMI-0 (they've sent us that instead of
> the double-DVI I wanted).
> 
> 
> It seems to use xf86-video-radeon:

In my experience that would be thing 0 to change.  I use radeonhd on my 
Radeon HD 4650, and though I've never used multiple screens on it, I have
had no end of problems with the radeon driver driving it (5.0_STABLE, amd64,
in case it matters).

Unfortunately, I have no help on the other issues,
--rafal

--

-- 
  Time is an illusion; lunchtime, doubly so.     |/\/\|           Rafal Boni
                   -- Ford Prefect               |\/\/|      rafal <at> pobox.com
Permalink | Reply | 0 comments |
headers
Henry Bent | 20 Feb 2010 03:22
Picon
Favicon

Intel X4500 (G41) and drm

Hi all,

I have an ASUS P5QPL-AM board using the Intel G41 chipset with Intel
X4500 onboard graphics.  At first i915drm did not recognize this chip,
but after adding the PCI ID (8086:2E32) to a few places in
sys/external/bsd/drm/dist and disabling drm's AGP support I was able
to get that driver to load correctly:

i915drm0 at vga0: Intel G41
i915drm0: Initialized i915 1.6.0 20080730

My problem now is that the xf86-video-intel driver seems to insist on
using AGP to set up this card, even though this motherboard doesn't
have any AGP support at all.  Loading drm is fine, but then it barfs
allocating memory.  /var/log/Xorg.0.log shows:

(EE) GARTInit: Unable to open /dev/agpgart (Device not configured)
(EE) intel(0): /dev/agpgart is either not available, or no memory is
available for allocation.  Please enable agpgart
.(WW) intel(0): VideoRam reduced to 32760 KB (page aligned - was 32763 KB)
(WW) intel(0): DRI2 requires UXA
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 9, (OK)
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 9, (OK)
drmOpenByBusid: Searching for BusID pci:0000:00:02.0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 9, (OK)
drmOpenByBusid: drmOpenMinor returns 9
drmOpenByBusid: drmGetBusid reports pci:0000:00:02.0
(Continue reading)

Permalink | Reply | 1 comment |
headers
Bernd Ernesti | 20 Feb 2010 08:52
Picon

Re: Intel X4500 (G41) and drm

On Fri, Feb 19, 2010 at 09:22:27PM -0500, Henry Bent wrote:
> Hi all,
> 
> I have an ASUS P5QPL-AM board using the Intel G41 chipset with Intel
> X4500 onboard graphics.  At first i915drm did not recognize this chip,
> but after adding the PCI ID (8086:2E32) to a few places in
> sys/external/bsd/drm/dist and disabling drm's AGP support I was able
> to get that driver to load correctly:

Hmm, what exactly did you do?

> i915drm0 at vga0: Intel G41
> i915drm0: Initialized i915 1.6.0 20080730
> 
[..]

> What am I doing wrong?  I have /dev/agpgart (linked to /dev/agp0) and
> agp support in my kernel, but the driver never loads ("dmesg | grep
> agp" shows nothing).

At least you need to add the G41 chip to src/sys/dev/pci/agp_i810.c

Bernd
Permalink | Reply | 0 comments |
headers
Michael Piotrowski | 22 Feb 2010 23:58
X-Face
Picon

Mouse behavior with mga(4) on 5.0.2

Hi,

I upgraded an i386 system from NetBSD 4.0.1 to 5.0.2.  The machine
contains a Matrox G550 AGP card, and I'm using its DVI output.

With XFree86 and its mga(4) driver, the mouse movement was very "choppy"
when using the DVI output.  Using the Matrox driver [1] fixed the
problem.

I see the same behavior with Xorg on 5.0.2.  The Matrox driver doesn't
work out of the box.  After some tinkering I found that doing

  # cd /lib
  # ln -s libc.so.12.164 libc.so.6

and starting X using something like

  # LD_LIBRARY_PATH=/lib /usr/X11R7/bin/startx -- -ignoreABI

does work and gives me back a usable mouse.

This is, however, a kludge.  I'm wondering whether the mouse issue with
mga(4) is a know problem and whether there is a better solution.  I've
put the dmesg output and the Xorg log at

  <http://dynalabs.de/mxp/netbsd/dmesg>
  <http://dynalabs.de/mxp/netbsd/Xorg.0.log>

All hints are appreciated!
(Continue reading)

Permalink | Reply | 2 comments |
headers
Zafer Aydogan | 25 Feb 2010 22:49
Picon

deadlocks while switching VTs

Hello list,

I'm experiencing some weird deadlocks while switching VTs
from or to X. It does not happen while switching between wscons terminals.
When the system deadlocks, all cpu's are used 100% and nothing works anymore.
I'm using modular-xorg from pkgsrc on 5.0.2. This happens on i386 and amd64.

Any ideas how to workaround this ?

Cheers, Zafer.
Permalink | Reply | 2 comments |
headers
Hubert Feyrer | 26 Feb 2010 00:11
Picon
Favicon

Re: deadlocks while switching VTs


Am 25.02.2010 um Do, 25. 02 2010, 22:49 schrieb Zafer Aydogan:
> I'm experiencing some weird deadlocks while switching VTs
> from or to X. It does not happen while switching between wscons  
> terminals.
> When the system deadlocks, all cpu's are used 100% and nothing works  
> anymore.
> I'm using modular-xorg from pkgsrc on 5.0.2. This happens on i386  
> and amd64.
>
> Any ideas how to workaround this ?

No workaround, but I think the generic way to debug this is:
try to boot with a serial console, and if things hang break into DDB,  
then see what the backtrace is.
And post it here.

  - Hubert
Permalink | Reply | 1 comment |
headers
Jean-Yves Migeon | 26 Feb 2010 01:25
Picon
Favicon

Re: deadlocks while switching VTs

On 02/26/10 00:11, Hubert Feyrer wrote:
>
> Am 25.02.2010 um Do, 25. 02 2010, 22:49 schrieb Zafer Aydogan:
>> I'm experiencing some weird deadlocks while switching VTs
>> from or to X. It does not happen while switching between wscons
>> terminals.
>> When the system deadlocks, all cpu's are used 100% and nothing works
>> anymore.
>> I'm using modular-xorg from pkgsrc on 5.0.2. This happens on i386 and
>> amd64.
>>
>> Any ideas how to workaround this ?
>
> No workaround, but I think the generic way to debug this is:
> try to boot with a serial console, and if things hang break into DDB,
> then see what the backtrace is.
> And post it here.

Add ps/l output (from within ddb too), to get info about the wait channels.

Cheers,

--

-- 
Jean-Yves Migeon
jeanyves.migeon <at> free.fr
Permalink | Reply | 0 comments |
headers
Jonathan A. Kollasch | 28 Feb 2010 21:50
Gravatar

Re: Mouse behavior with mga(4) on 5.0.2

On Mon, Feb 22, 2010 at 11:58:47PM +0100, Michael Piotrowski wrote:
> Hi,
> 
> I upgraded an i386 system from NetBSD 4.0.1 to 5.0.2.  The machine
> contains a Matrox G550 AGP card, and I'm using its DVI output.
> 
> With XFree86 and its mga(4) driver, the mouse movement was very "choppy"
> when using the DVI output.  Using the Matrox driver [1] fixed the
> problem.

Doesn't sound unexpected given the blurb in mga(4):

[That module also provides various other enhancements,
and may be necessary to use the DVI output on the G550.]

	Jonathan Kollasch.
Permalink | Reply | 1 comment |
headers
Michael Piotrowski | 1 Mar 2010 16:15
X-Face
Picon

Re: Mouse behavior with mga(4) on 5.0.2

On 2010-02-28, "Jonathan A. Kollasch" <jakllsch <at> kollasch.net> wrote:

>> I upgraded an i386 system from NetBSD 4.0.1 to 5.0.2.  The machine
>> contains a Matrox G550 AGP card, and I'm using its DVI output.
>> 
>> With XFree86 and its mga(4) driver, the mouse movement was very "choppy"
>> when using the DVI output.  Using the Matrox driver [1] fixed the
>> problem.
>
> Doesn't sound unexpected given the blurb in mga(4):
>
> [That module also provides various other enhancements,
> and may be necessary to use the DVI output on the G550.]

Well yes, but "may be necessary" is not the same as "is required" :-) In
fact the DVI output works fine--it's the mouse that doesn't, and mga(4)
doesn't say anything about that.

Essentially I was wondering whether others have also experienced this
issue since I couldn't find anything about it on the Web.

Best regards

--

-- 
Dr.-Ing. Michael Piotrowski, M.A.                   <mxp <at> dynalabs.de>
Public key at <http://www.dynalabs.de/mxp/pubkey.txt> (ID 0x1614A044)
Permalink | Reply | 0 comments |

Gmane